Risk factors for intensive care unit delirium after cardiac operation

Abstract

To analyze the risk factors of delirium in patients in cardiac surgery intensive care unit (CSICU). A prospective observational study was performed. Patients admitted to CSICU of Fujian Medical University Union Hospital from March to August in 2017 were enrolled. The combination of the Richmond agitation sedation scale (RASS) and the ICU-confusion assessment method (CAM-ICU) were used to evaluate delirium. The patient was assessed on the second day after CSICU admission, twice a day, the evaluation was stopped, and the follow-up observation was terminated after the patient was discharged from CSICU. The patients were divided into two groups according to whether delirium occurred in CSICU. The general and clinical treatment data (including condition, operation, anesthesia and CSICU treatment) of the two groups were compared. The related factors of delirium were identified by univariate analysis and multifactor Logistic regression analysis. A total of 318 cases were included in this study. Among them, 93 cases had delirium and the incidence of delirium was 29.2%. It was shown by univariate analysis that age, history of hypertension, type of surgery, surgical procedure, American Society of Anesthesiologists (ASA) anesthesia classification, usage of propofol, plasma transfusion, red blood cells, platelet transfusion, blood loss, operative time, cardiopulmonary bypass (CPB) time, myocardial block time, acute physiology and chronic health evaluation II (APACHE II), duration of mechanical ventilation, the length of intensive care unit (ICU) stay, postoperative usage of diazepam, midazolam, fentanyl, morphine, chlorpromazine, etc. which were related to delirium, and occupation (on-the-job or self-employed), medical insurance (city or provincial medical insurance), education (primary to junior high school, high school or above) could reduce the risk of delirium. Colinearity diagnosis was performed on variables with statistically significant differences, and variables with variance expansion factor (VIF) < 3 were included in multivariate Logistic regression analysis. The results showed that age, education level, type of surgery, ASA classification, CPB time, APACHE II, ICU mechanical ventilation time, and post operation usage of midazolam were independently related to delirium [age: odds ratio (OR) = 1.625, 95% confidence interval (95%CI) = 1.303-2.026; education level: OR = 0.293, 95%CI = 0.171-0.504; type of surgery: OR = 2.194, 95%CI = 1.052-4.576; ASA classification: OR = 1.916, 95%CI = 1.032-3.559; CPB time: OR = 2.125, 95%CI = 1.105-4.088; APACHE II: OR = 2.091, 95%CI = 1.005-4.349; ICU mechanical ventilation time: OR = 1.943, 95%CI = 1.269-2.975; midazolam: OR = 2.653, 95%CI = 1.328-5.299; all P < 0.05], among which, high education level has a good protective effect on delirium. Age, type of surgery, ASA classification, CPB time, APACHE II, ICU mechanical ventilation time, post operation usage of midazolam were independent risk factors for delirium, and high education level had a good protective effect. Among them, the educational level, CPB time, duration of mechanical ventilation, and midazolam are intervention factors. In clinical treatment, not only the risk factors should be identified, but also intervention should be taken to prevent the occurrence of delirium.